Collaborative Research: Low Peak to Average Power Multicarrier Signals via Coding: Fundamental Limits and Algorithms

在部署高数据速率、低成本和功率高效的多载波通信系统中，最重要的障碍之一是前端RF放大器的成本。在认知网络中，多载波调制提供了低复杂度均衡、对信道频率选择性的适应性以及有效利用可用带宽的承诺。然而，多载波信号由许多子载波的线性叠加组成，导致大的峰均功率比（PAPR），这意味着需要高度线性的功率放大器。特别地，当RF功率放大器的线性区域增加以适应具有大PAPR的信号时，RF功率放大器变得基本上功率低效且昂贵。由功率放大器引入的任何非线性都可能导致大的带外泄漏并减小链路的传输范围。虽然最坏情况下的PAPR相当大，但研究人员已经确定，具有大峰值的可能性很小，并且存在PAPR受常数限制的几乎全速率的代码。然而，现有的编码方案提供了低PAPR和大的最小距离，其代价是大大降低了传输速率。这项研究涉及解决存在的和算法方面的编码方案的PAPR降低。本研究开发了概率技术和新的算法来证明存在性，并构造实用的代码，不仅提供低PAPR和高速率，但也导致优化其他参数，如容量损失，最小距离和带外泄漏。

One of the most important obstacles in deployment of high data rate, low cost, and power efficient multicarrier communication systems is the cost of front-end RF amplifiers. Multicarrier modulations offer the promise of low complexity equalization, adaptability to frequency selectivity of the channel, and efficient use of available bandwidth in cognitive networks. However, multicarrier signals consist of linear superposition of many subcarriers leading to large peak to average power ratio (PAPR) and implying the need for highly linear power amplifier. In particular, RF power amplifier becomes substantially power inefficient and expensive when its linearity region increases to accommodate signals with large PAPR. Any nonlinearity introduced by the power amplifier can cause large out-of-band leakage and reduce the transmit range of the link.While the worst case PAPR is quite large, it has been established by the investigators that the likelihood of having large peaks is small and there exist codes of almost full rate with PAPR bounded by a constant. Existing coding schemes, however, provide low PAPR and large minimum distance at the cost of substantially reducing the transmission rate. This research involves addressing both existential and algorithmic aspects of coding schemes for PAPR reduction. This research develops both probabilistic techniques and new algorithms to prove the existence and to construct pragmatic codes that not only provide low PAPR and high rate but also lead to optimizing other parameters such as capacity loss, minimum distance and out-of-band leakage.